Project Info 
(Completed)

SoftEdge: Architectures and Algorithms for Software Defined Edge Systems

Project 17/CDA/4760, Career Development Award, 2018-2021; Single-PI project. 

 

Sponsored by Science Foundation Ireland​; Partners: Yale University; IBM T. J. Watson Research Center(NY); NEC Research Labs (Heidelberg); CONNECT (Dublin)

Today we are witnessing the impressive transformation of the wireless networks. The demand has dramatically shifted from plain voice to throughput-hungry and low-latency communication and computing services, and new types of users have emerged. In order to address these challenges the emerging 5G+ systems include several innovations, two of the most promising being: (i) leverage the SDN/NFV technologies and create fully-programmable wireless systems, and (ii) push services closer to demand by harnessing resources at the network edge. SoftEdge lies at the nexus of these game-changing ideas and aspires to unleash their potential by addressing key bottleneck issues. Namely, it will propose a new class of multi-tier softwarized edge systems; introduce a rigorous methodology for designing their architecture and control fabric; and deliver novel resource orchestration policies for supporting demanding services such as mobile data analytics. Moreover, this project will study the economics of edge systems and design virtualization and pricing mechanisms that will maximize their efficiency and performance. SoftEdge has been carefully designed to maximize impact: it brings together world-renowned academics and industry labs; includes both theoretical and experimental tasks; evaluates representative use cases and demos; has identified paths to commercialization; and has a set of innovative education and outreach activities

Project Info
(Completed)

SERHENA: Service-Aware Resource Orchestration in Hybrid Cloud/Edge  Architectures

Marie-Sklodowska Curie Actions, co-sponsored by H2020, EDGE COFUND Fellowship, 2017 - 2020.

Hosted  Researcher: Dr. Jungho Kwak

A very promising solution for addressing the increasing network congestion problem is edge-networking: the deployment and smart management of network, computing and storage resources at the network edge. This approach can substantially improve the services offered to users, and reduce the required network expenditures. To this end, SERHENA will deliver a multi-faceted optimization framework aiming to unleash the full potential of edge networking by addressing certain bottleneck issues.

 

SERHENA incorporates many innovations, as it will: optimize the resource dimensioning in hybrid EN systems that include both central and edge resources; jointly orchestrate different types of resources (e.g., computing and bandwidth) so as to satisfy the emerging types of services; study methods to leverage idle user-owned equipment; propose online resource management policies that achieve asymptotic cooperation equilibriums; focus on service-aware network optimization techniques; and analyze the market dynamics among the service providers and operators in this new era so as to create sustainable solutions. SERHENA will follow a novel methodology, using both rigorous analytical tools, and data-driven or testbed-based evaluation approaches. 

Project Info
(Completed)

Video Delivery Optimization for Next Generation Wireless Networks
Targeted Industry Project, Co-sponsored by SFI/CONNECT and Nokia Ireland, 2017-2020.

In this project we design, implement and evaluate algorithms for revolutionizing the next generation mobile video delivery services, using an innovative mix of novel architectures and optimization algorithms.

Project Info
(Ongoing)

Network intelligence for aDAptive and sElf-Learning MObile Networks
Horizon 2020, 2021-2023. [Project website]

The DAEMON H2020 European project develops and implements innovative and pragmatic approaches to Network Intelligence (NI) design that enable high performance, sustainable and extremely reliable zero-touch network system. DAEMON designs an end-to-end NI-native architecture for Beyond 5G (B5G) that fully coordinates NI-assisted functionalities.